arXiv subjects

Cunzhi Zhao

Publications and source records attributed to Cunzhi Zhao.

14 recordsLinked to original sources

arXiv2026

Economic Evaluation of V2G-Enabled Fast Charging Stations Under Endogenous EV Adoption Dynamics

Building fast charging stations (FCSs) is crucial for transportation electrification, but there exists an indirect network effect: while the increasing number of electric vehicles (EVs) decides the FCS capacity expansion, the spatial locations of these facilities strongly influence drivers' willingness to adopt EVs. Ignoring this interaction can lead to bad capital investments and exacerbate power grid vulnerabilities during tidal traffic peaks. Therefore, we explicitly model the EV adoption dynamics as decision-dependent uncertainties (DDUs) in a new multi-period collaborative planning framework. This framework evaluates the economic viability of V2G-enabled FCSs across both transportation and distribution networks. To simplify the complex calculation, we introduce an aggregated fleet virtual battery model to catch macroscopic vehicle-to-grid (V2G) flexibility. This successfully circumvents the dimension curse inherent in tracking microscopic state-of-charge. To further guarantee calculation speed, the nonlinear infrastructure exposure is transformed into a mixed-integer program by using Special Ordered Set type 2 (SOS2) variables and Second-Order Cone Programming (SOCP) relaxations for grid limits. Finally, numerical studies on a coupled Sioux Falls and IEEE 33-bus testbed prove that our framework achieves superior expected social welfare. Also, macroscopic V2G aggregation is highlighted for its capability to mitigate distribution grid congestion penalties.

eess.SY↗

arXiv2025

Optimal Power Scheduling for High Renewables-Integrated Energy Systems with Battery Storage

In high renewables-integrated power systems, irrespective to their sizes, energy storage is commonly included and utilized to mitigate fluctuations from both the load and renewable power generation, ensuring system reliability, among which battery energy storage system (BESS) are experiencing fast-growth in recent years. The BESS systems, predominantly employing lithi-um-ion batteries, have been extensively deployed. The degrada-tion of these batteries significantly affects system efficiency. Deep neural networks can accurately quantify the battery degrada-tion; however, the model complexity hinders their applications in energy scheduling for various power systems at different scales. To address this issue, this paper presents a novel approach, in-troducing a linearized sparse neural network-based battery deg-radation model (SNNBD), specifically tailored to quantify battery degradation based on the scheduled BESS operational profiles. This approach achieves accurate degradation prediction while substantially reducing the complexity associated with a dense neural network model. The computational burden of day-ahead energy scheduling when integrating battery degradation can thus be substantially alleviated. Case studies, conducted on both small-scale microgrids and large-scale bulk power grids, demonstrated the efficiency and suitability of the proposed optimal energy scheduling model that can effectively address battery degrada-tion concerns while optimizing day-ahead energy scheduling op-erations.

eess.SY↗

arXiv2025

Linearization of ReLU Activation Function for Neural Network-Embedded Optimization: Optimal Day-Ahead Energy Scheduling

Recently, neural networks have been widely applied in the power system area. They can be used for better predicting input information and modeling system performance with increased accuracy. In some applications such as battery degradation neural network-based microgrid day-ahead energy scheduling, the input features of the trained learning model are variables to be solved in optimization models that enforce limits on the output of the same learning model. This will create a neural network-embedded optimization problem; the use of nonlinear activation functions in the neural network will make such problems extremely hard to solve if not unsolvable. To address this emerging challenge, this paper investigated different methods for linearizing the nonlinear activation functions with a particular focus on the widely used rectified linear unit (ReLU) function. Four linearization methods tailored for the ReLU activation function are developed, analyzed and compared in this paper. Each method employs a set of linear constraints to replace the ReLU function, effectively linearizing the optimization problem, which can overcome the computational challenges associated with the nonlinearity of the neural network model. These proposed linearization methods provide valuable tools for effectively solving optimization problems that integrate neural network models with ReLU activation functions

cs.LG↗

arXiv2024

Comparative Analysis of Machine Learning Models for Short-Term Distribution System Load Forecasting

Accurate electrical load forecasting is crucial for optimizing power system operations, planning, and management. As power systems become increasingly complex, traditional forecasting methods may fail to capture the intricate patterns and dependencies within load data. Machine learning (ML) techniques have emerged as powerful alternatives, offering superior prediction accuracy and the ability to model non-linear and complex temporal relationships. This study presents a comprehensive comparison of prominent ML models: feedforward neural networks, recurrent neural networks, long short-term memory networks, gated recurrent units, and the attention temporal graph convolutional network; for short-term load forecasting of the Energy Corridor distribution system in Houston, Texas. Using a 24-hour look-back window, we train the models on datasets spanning one and five years, to predict the load demand for the next hour and assess performance. Our findings aim to identify the most effective ML approach for accurate load forecasting, contributing to improved grid reliability and system optimization.

eess.SY↗

arXiv2024

Hierarchical Deep Learning Model for Degradation Prediction per Look-Ahead Scheduled Battery Usage Profile

Batteries can effectively improve the security of energy systems and mitigate climate change by facilitating wind and solar power. The installed capacity of battery energy storage system (BESS), mainly the lithium ion batteries are increasing significantly in recent years. However, the battery degradation cannot be accurately quantified and integrated into energy management system with existing heuristic battery degradation models. This paper proposed a hierarchical deep learning based battery degradation quantification (HDL-BDQ) model to quantify the battery degradation given scheduled BESS daily operations. Particularly, two sequential and cohesive deep neural networks are proposed to accurately estimate the degree of degradation using inputs of battery operational profiles and it can significantly outperform existing fixed or linear rate based degradation models as well as single-stage deep neural models. Training results show the high accuracy of the proposed system. Moreover, a learning and optimization decoupled algorithm is implemented to strategically take advantage of the proposed HDL-BDQ model in optimization-based look-ahead scheduling (LAS) problems. Case studies demonstrate the effectiveness of the proposed HDL-BDQ model in LAS of a microgrid testbed.

eess.SY↗

arXiv2023

Microgrid Optimal Energy Scheduling with Risk Analysis

Risk analysis is currently not quantified in microgrid resource scheduling optimization. This paper conducts a conditional value at risk (cVaR) analysis on a grid-disconnected residential microgrid with distributed energy resources (DER). We assume the infrastructure to set up an ad-hoc microgrid is already in place for a residential neighborhood with power sources such as photovoltaic (PV), diesel, and battery energy storage system (BESS). With this scenario in mind, we solve day-ahead scheduling to optimally allocate various resources to match demand in scenarios where neighborhoods, especially residential, are disconnected from the overall grid such as in flooding, hurricanes, winter storms, or operational failures. The goal is to provide an alternative framework to optimize power availability for priority customers and strengthen the overall grid against dips in power outside of normal operating considerations. The focus of this paper will be taking in renewable energy sources from PV combined with diesel and BESS while minimizing cost. Case studies demonstrate that with the proposed energy management system, microgrids can be implemented to be more resilient against new challenges.

eess.SY↗

arXiv2022

Microgrid Optimal Energy Scheduling Considering Neural Network based Battery Degradation

Battery energy storage system (BESS) can effec-tively mitigate the uncertainty of variable renewable generation. Degradation is unpreventable and hard to model and predict for batteries such as the most popular Lithium-ion battery (LiB). In this paper, we propose a data driven method to predict the bat-tery degradation per a given scheduled battery operational pro-file. Particularly, a neural network based battery degradation (NNBD) model is proposed to quantify the battery degradation with inputs of major battery degradation factors. When incorpo-rating the proposed NNBD model into microgrid day-ahead scheduling (MDS), we can establish a battery degradation based MDS (BDMDS) model that can consider the equivalent battery degradation cost precisely with the proposed cycle based battery usage processing (CBUP) method for the NNBD model. Since the proposed NNBD model is highly non-linear and non-convex, BDMDS would be very hard to solve. To address this issue, a neural network and optimization decoupled heuristic (NNODH) algorithm is proposed in this paper to effectively solve this neural network embedded optimization problem. Simulation results demonstrate that the proposed NNODH algorithm is able to ob-tain the optimal solution with lowest total cost including normal operation cost and battery degradation cost.

eess.SP↗

arXiv2022

Quality Analysis of Battery Degradation Models with Real Battery Aging Experiment Data

The installation capacity of energy storage system, especially the battery energy storage system (BESS), has increased significantly in recent years, which is mainly applied to mitigate the fluctuation caused by renewable energy sources (RES) due to the fast response and high round-trip energy efficiency of BESS. The main components of majority of BESSs are lithium-ion batteries, which will degrade during the BESS daily operation. Heuristic battery degradation models are proposed to consider the battery degradation in the operations of energy systems to optimize the scheduling. However, those heuristic models are not evaluated or demonstrated with real battery degradation data. Thus, this paper will perform a quality analysis on the popular heuristic battery degradation models using the real battery aging experiment data to evaluate their performance. A benchmark model is also proposed to represent the real battery degradation value based on the averaged cycle value of the experimental data.

eess.SY↗

arXiv2022

A 100% Renewable Energy System: Enabling Zero CO2 Emission Offshore Platforms

The total electricity consumption from offshore oil/gas platforms is around 16 TWh worldwide in 2019. The majority offshore platforms are powered by the diesel generators while the rest mainly uses gas turbines, which emits large amounts of CO2 per year. The fast development of offshore wind turbines (WT) can potentially replace traditional fossil fuel based resources to power offshore loads. Thus, a novel offshore hybrid renewable energy sources (OHRES) system is proposed to enable a zero CO2 emission offshore platform mitigating climate change. Battery energy storage system (BESS) and hydrogen energy storage system (HESS) are considered to mitigate the fluctuation of wind power in the proposed OHRES system. Resilience models are designed to enhance the resilience of the proposed OHRES system with extra energy stored in BESS and/or HESS. Case studies demonstrate the feasibility of the proposed OHRES system to power offshore platforms. The economic analysis reports the planning cost for the proposed OHRES system under different resilience levels, which may benefit the decision to balance the carbon emission and investment cost.

eess.SY↗

arXiv2022

An Alternative Method for Solving Security-Constrained Unit Commitment with Neural Network Based Battery Degradation Model

Battery energy storage system (BESS) can effectively mitigate the uncertainty of variable renewable generation and provide flexible ancillary services. However, degradation is a key concern for rechargeable batteries such as the most widely used Lithium-ion battery. A neural network based battery degradation (NNBD) model can accurately quantify the battery degradation. When incorporating the NNBD model into security-constrained unit commitment (SCUC), we can establish a battery degradation based SCUC (BD-SCUC) model that can consider the equivalent battery degradation cost precisely. However, the BD-SCUC may not be solved directly due to high non-linearity of the NNBD model. To address this issue, the NNBD model is linearized by converting the nonlinear activation function at each neuron into linear constraints, which enables BD-SCUC to become a linearized BD-SCUC (L-BD-SCUC) model. Case studies demonstrate the proposed L-BD-SCUC model can be efficiently solved for multiple BESS buses power system day-ahead scheduling problems with the lowest total cost including the equivalent degradation cost and normal operation cost.

eess.SY↗

arXiv2022

Resilient Operational Planning for Microgrids Against Extreme Events

This paper proposes a novel resilience index, a microgrid survivability rate (SR) under extreme events, and then proposes a novel Resilient Operational Planning (ROP) algorithm to maximize the proposed resilience index SR. The proposed ROP algorithm can incorporate predetermined inverter failure probabilities and generate multiple scenarios accordingly to optimize resilient operations during an extreme event. The implemented ROP algorithm consists of two main steps: (i) optimization of resilient operational planning, and (ii) preventive resilience enhancement if minimum SR is not met per the analysis in step 1. A typical microgrid (MG) is studied to compare the proposed ROP algorithm against a traditional microgrid energy management (MEM) model. Results indicate that an enhanced resilience operation is achieved by the ROP algorithm, which is demonstrated by the quantification of resilience via the SR. Moreover, the proposed ROP algorithm is able to obtain a greater SR overall compared to that achieved by the traditional MEM, and this benefit of using the proposed ROP increases as the inverter failure probabilities increase.

eess.SY↗

arXiv2021

Quantitative Analysis of Demand Response Using Thermostatically Controlled Loads

The flexible power consumption feature of thermostatically controlled loads (TCLs) such as heating, ventilation, and air-conditioning (HVAC) systems makes them attractive targets for demand response (DR). TCLs possess a brief period where their power utilization can be altered without any significant impact on customer comfort level. This indicates TCLs are hidden potentials for providing ancillary services. This paper proposes a novel metric of demand response support time (DRST) for HVAC enabled demand response and a novel algorithm for the quantification of such HVAC-DR. The consumers' comfort will not be compromised with the proposed DRST-based HVAC-DR. Case studies demonstrate its benefits in terms of cost saving in microgrid day-ahead scheduling and reduction of forced load shedding during a grid-microgrid tie-line outage event. This illustrates the reserve potential benefits and the increase of microgrid reliability when DRST-based HVAC-DR is considered.

eess.SY↗

arXiv2021

A Novel Real-Time Energy Management Strategy for Grid-Supporting Microgrid: Enabling Flexible Trading Power

In recent years, there has been significant growth of distributed energy resources (DERs) penetration in the power grid. The stochastic and intermittent features of variable DERs such as roof top photovoltaic (PV) bring substantial uncertainties to the grid on the consumer end and weaken the grid reliability. In addition, the fact that numerous DERs are widespread in the grid makes it hard to monitor and manage DERs. To address this challenge, this paper proposes a novel real-time grid-supporting energy management (GSEM) strategy for grid-supporting microgrid (MG). This strategy can not only properly manage DERs in a MG but also enable DERs to provide grid services, which enables a MG to be grid-supporting via flexible trading power. The proposed GSEM strategy is based on a 2-step optimization which includes a routine economic dispatch (ED) step and an acceptable trading power range determination step. Numerical simulations demonstrate the performance of the proposed GSEM strategy which enables the grid operator to have a dispatch choice of trading power with MG and enhance the reliability and resilience of the main grid.

eess.SY↗

arXiv2020

A Novel Real-Time Energy Management Strategy for Gird-Friendly Microgrid: Harnessing Internal Fluctuation Internally

Typically, a large portion of microgrid generating capacity is from variable renewable resources that are greatly impacted by the environment and can be intermittent as well as stochastic. This would result in uncertainty of microgrid net-load, and negatively affect the grid reliability. A two-phase real-time energy management strategy for networked microgrid is proposed in this paper to address microgrid internal fluctuation internally, which enables a microgrid to become grid-friendly. The proposed strategy is based on coordination between the real-time dispatch (RTD) phase and the real-time control (RTC) phase. In the RTD phase, model predictive control (MPC) is used to optimally dispatch microgrid resources in the current time interval while considering near future situations. The RTC phase addresses microgrid internal net-load fluctuation with fast-acting batteries, which aims to maintain a constant tie-line power flow between the main grid and the microgrid for the current dispatch interval. Numerical simulations conducted on ten different net-load scenarios can demonstrate the performance of the proposed two-phase energy management strategy that will enable a microgrid to operate as a controllable asset with static electricity consumption or production in an economic dispatch interval from the perspective of the bulk grid operator.

eess.SY↗